Journal of Behavior Therapy

and Experimental Psychiatry 30 (1999) 191}198

Reducing test anxiety and improving academic

self-esteem in high school and college students
with learning disabilities
Donald Wachelka, Roger C. Katz*
Department of Psychology, University of the Pacixc, Stockton, CA 95211, USA

Abstract

Test anxiety seems like a benign problem to some people, but it can be potentially serious
when it leads to high levels of distress and academic failure in otherwise capable students.
Because test anxiety is common in older students with learning disabilities (LD), it is surprising
that little research has been done on ways to reduce the distress these students experience in test
situations. In this study, we used a randomized pretest}posttest control group design to
examine the e!ectiveness of a cognitive-behavioral treatment for reducing test anxiety and
improving academic self-esteem in a cohort (N"27) of high school and college students with
learning disabilities (LD). All of the students participated voluntarily. They were enrolled in
classes for students with learning problems. Before the study began, they complained of test
anxiety and showed an elevated score on the Test Anxiety Inventory (TAI). Eleven students
(85%) completed the 8-week long treatment, which consisted of progressive muscle relaxation,
guided imagery, self-instruction training, as well as training in study and test-taking skills.
Results showed signi"cant improvement in the treated group which was not evident in an
untreated control group (N"16). Compared to the control group, the treated group showed
signi"cant reductions in test anxiety on the TAI, as well as improvement in study skills and
academic self-esteem as measured by the Survey of Study Habits and Attitudes, and the school
scale of the Coopersmith Self-Esteem Inventory. These results extend the generality of similar
studies on reducing test anxiety and improving academic self-esteem in younger students. They
also suggest that relief from test anxiety can be expected fairly quickly when cognitive-
behavioral methods are used. Additional implications and methodological limitations of the
study are discussed. g 1999 Published by Elsevier Science Ltd. All rights reserved.

Keywords: Test anxiety; Learning disabilities

* Corresponding author.
E-mail address: [email protected] (R.C. Katz)

0005-7916/99/$ - see front matter g 1999 Published by Elsevier Science Ltd. All rights reserved.
PII: S 0 0 0 5 - 7 9 1 6 ( 9 9 ) 0 0 0 2 4 - 5
192 D. Wachelka, R.C. Katz / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 191}198

1. Introduction

Test anxiety involves the unpleasant experience of worry and emotionality in

situations where the person feels he or she is being evaluated (Dusek, 1980). The
evaluation of academic achievement by tests is a prime example.
The e!ects of test anxiety are mediated by several variables including the level of
anxiety, the di$culty of the task, and the ability of the student (Ball, 1995). A small
amount of anxiety acts as a motivator; it can enhance performance by encouraging the
student to try. Conversely, too much anxiety has the opposite e!ect: it can disrupt
mental processes that are needed for the student to perform well. This is especially true
when the learning task is demanding, as it often is in the assessment of academic
achievement.
Among high school and college students, test anxiety is a common and potentially
serious problem. Debilitating test anxiety a!ects 10}30% of all students, with a dis-
proportionately higher prevalence in learning disabled and minority students (Bryan,
Sonnefeld & Grabowski, 1983; Nicaise, 1995; Strumph & Fodor, 1993). Twenty-
percent of test anxious students quit school before graduating because of repeated
academic failure (Tobias, 1979). High test anxiety is also associated with low self
esteem, poor reading and math achievement, failing grades, disruptive classroom
behavior, negative attitudes toward school, and unpleasant feelings of nervousness
and dread that stem from an intense fear of failure (Bryan et al., 1983). Because test
anxiety has many adverse e!ects on the lives of students, and the accurate assessment
of their academic achievement, it is an important topic for research.
While e!ective methods for treating test-taking anxiety have been described pre-
viously (Nicaise, 1995; Sapp, 1994; Strumph & Fodor, 1993), these studies typically
involved non-handicapped students. In our review of the literature, we found no
controlled studies on the treatment of test anxiety in high school and college students
with learning disabilities, despite the increased prevalence of test anxiety in this
population and the fact that they are already at greater risk for academic failure
because of their skill de"ciencies (Rizzo & Zabel, 1988; Scruggs & Mastropieri,
1988; Swanson & Howell, 1996). Academic success and advancement at all
levels is based on test scores; therefore, it is crucial that students with learning
handicaps know how to prepare for and take tests without feeling overwhelmed by
anxiety.
In this study, we describe a cognitive-behavioral treatment for reducing test taking
anxiety and improving academic self-esteem in high school and college students with
learning disabilities. The treatment approach incorporated several techniques that
showed promise in previous studies (Nicaise, 1995; Sapp & Farrell, 1994; Vagg
& Speilberger, 1995). The intervention consisted of relaxation training, guided im-
agery, and self-instructional training for reducing increased arousal and stress-
heightening cognitions, and study skills training to improve e!orts to prepare for and
take examinations. Based on previous research (e.g., Sapp & Farrell, 1994; Wessel
& Mersch, 1994), we expected that students with LD who received treatment would
report less anxiety and more con"dence in test taking situations than comparable
students who did not receive treatment.
 D. Wachelka, R.C. Katz / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 191}198 193

2. Method

2.1. Participants

Twenty-nine public high school and junior college students (14 males; 15 females)
volunteered to participate. All of the students were diagnosed as having a LD by
school personnel which meant that they had normal intelligence but displayed
signi"cant di$culties in the area of reading writing, or arithmetic. All participants
reported that test anxiety was a problem for them.¹ Twenty-eight of the students were
recruited from a resource class for students with LD. The remaining participant was
a high school student who was referred from a private clinic that provided services to
students with learning problems. The students ranged in age from 17 to 52 years old.
They were all Caucasian. Their average age was 28.72 (SD"12.65) years. Of the 27
students who completed the study, 15 were in high school and 12 were in junior
college. Prospective subjects were told that the purpose of the study was to reduce
their anxiety about taking tests and improve their attitude toward schoolwork. They
were o!ered no other incentive for their participation. Informed consent was obtained
before the study began.

2.2. Design

The study was designed as a randomized pretest}posttest control group experi-

ment. Thirteen students were randomly assigned to the treatment group; 16 students
were randomly assigned to a waiting list control group. The groups were equivalent in
terms of age and gender composition.* Two of the students in the treatment group
dropped out shortly after the study began, one because of marital problems and the
other because of school attendance problems. This left a total of 11 students (85%)
who completed the 8 week long treatment.

2.3. Measures

Three dependent measures were used: the Test Anxiety Inventory (TAI) (Speil-
berger, 1980), the study orientation (SO) sub-scale of the Survey of Study Habits and
Attitudes-Form H (Brown & Holtzman, 1967), and the school scale (SS) of the

¹ A screening device was used to determine self-reported di$culty with test situations. The questions used
were as follows: When taking tests do you often (a) Experience extreme bodily tension? (b) Worry about
whether you will pass? (c) Go blank, even when you know the answers? (d) Feel hurried, inadequate, or
panicked and experience lessened self-esteem? If the student said yes to any of these questions, he/she was
eligible to participate in an initial intake which included an interview and preliminary assessment. If the
student scored at least a 40 on the TAI, he/she was eligible to participate in the study.
* The treatment group consisted of 5 males and 6 females whose mean age as 25.73 years (SD 10.37).
Conversely, the control group consisted of 9 males and 7 females whose mean age was 30.31 years
(SD"13.71). Neither of these di!erences was signi"cant.
194 D. Wachelka, R.C. Katz / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 191}198

Coopersmith Self-Esteem Inventory (Coopersmith, 1981). Each of the measures has

acceptable reliability and validity, and has been used in previous studies of this nature
(Sapp, 1994; Swanson & Howell, 1996).
The TAI is a 20-item self-report measure that uses a Likert format to assess anxiety
and discomfort in test-taking situations. Sample items include, `I feel very panicky
when I take an important testa and `I worry a great deal before taking an important
examination.a Scores on the TAI can range from 20 to 80. Higher scores indicate
higher levels of test anxiety.
The study orientation sub-scale of the Survey of Study Habits and Attitudes also
uses a Likert format to assess test-taking and study skills, as well as overall attitudes
about school work. On this measure, students were asked to rate 100 items according
to how true they were for them. Their answers could range from rarely true to almost
always true. Sample items include, `In preparing reports or term papers, I make
certain that I clearly understand what is wanted before I begin the worka, `Even
though I do not like a subject, I still work hard to make a good gradea, and `My
dislike for certain teachers causes me to neglect my school work.a Scores on the SO
can range from 0 to 200, with higher scores indicating better study habits and more
positive attitudes toward school.
The school sub-scale of the Coopersmith Self-Esteem Inventory measures academic
self-esteem. Students were asked to rate eight items as either true or false for them.
Items include, `I often feel upset in schoola, `I am not doing as well in school as
I would like toa, and `I often get discouraged at school.a Scores on the SS can range
from 0 to 8. Higher scores re#ect greater self-esteem and con"dence about school-
work.

2.4. Procedure

All of the students were pre-tested individually. Those in the treatment group were
then scheduled for treatment, which was provided individually during sessions that
were held 1 hour per week over 8 consecutive weeks. Treatment was provided by the
"rst author, a master's student in psychology. Students in the waiting list control
group received no treatment. They were told that the current treatment group was full
but that another class would begin at a later date. Treatment³ was given in the
following sequence:

Weeks 1}2: Students were instructed in and practiced progressive muscle relaxation
(PMR) (Goldfried & Davison, 1994). They were given a rationale for PMR * i.e.,
relaxation reduces tension and is incompatible with anxiety. During the "rst session,
they listened to a relaxation tape (Surwit, 1983) and practiced it under the direction of
the therapist. A copy of the tape was made available to the students for home practice

³ A treatment manual that describes the entire intervention in more detail is available upon request.
Correspondence should be addressed to Dr. Katz.
 D. Wachelka, R.C. Katz / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 191}198 195

which they were encouraged to do at least four times a week. They were also trained in
an abbreviated form of PMR for developing deep muscle relaxation quickly.
Week 3}4: Students were trained to recognize and dispute irrational beliefs, and to
engage in more rational self-talk instead (Ellis, 1985). They were told that negative
beliefs give rise to negative emotions, so if they changed their negative attitudes about
tests, the negative feelings that go along with them should gradually diminish. They
were given the following rationale for disputing negative beliefs: (a) People respond
di!erently to similar situations. (b) The way people perceive events a!ects how they
feel about them. (c) If people can change their negative attitudes, the negative feelings
associated with them will gradually weaken. During this phase, students were intro-
duced to concepts like `awfulizinga and `absolutizinga, and asked how these concepts
applied to their own situation. Students also completed a Beliefs Inventory (Ellis, as
cited in Davis, Eshelman & McKay, 1996) to identify speci"c irrational beliefs that
might be contributing to their distress. Once this was done, they were taught
a "ve-step plan for changing these beliefs. The plan consisted of the following steps:
(a) Describe the situation by focusing on factual events. (b) Describe a rational
response to the situation. (c) Pay attention to the relationship between your thoughts
(what you say to yourself) and feelings. (d) When necessary, practice rational dis-
puting. (e) Replace irrational self-talk with rational self-talk.
Week 5}6: Students were instructed in, and given a rationale for, guided imagery
(Sapp, 1994). As described by Goldfried and Davison (1994), the guided imagery was
incorporated into a self-systematic desensitization procedure where the students
constructed a hierarchy of anxiety producing events, in this case involving test-taking
situations such as having an examination announced in class, studying for an exam, or
taking an exam. In doing the self-desensitization, the students were taught to imagine
scenes from the hierarchy as vividly as possible by incorporating relevant stimuli
(sights, sounds, etc.) into the image. During these sessions, the students gradually
worked through the hierarchy while they were relaxed and engaged in positive self-
talk. They were also encouraged to practice guided imagery at home each day for a
week.
Week 7}8: These sessions focused on strategies for exam preparation. Students were
told that poor preparation contributes to anxiety in test-taking situations. They were
taught to manage their review time more e!ectively, use review tools such as study
checklists and #ashcards, and to conceptualize the topic they were learning about as
an outline that progresses from the general to the speci"c (Sapp, 1993). During this
phase, students were encouraged to monitor their study time and record the amount
of time spent studying for each exam. They were also given suggestions for improving
their study behavior (Sapp & Farrell, 1994). Among the suggestions presented
were the following: (a) Set aside a certain amount of time each day to study.
(b) Associate the completion of study assignments with enjoyable activities. Once
speci"c assignments are completed, reward yourself. (c) Study alone. Study in a group
only to review material already learned. (d) Study one subject for at least 45 min.
Avoid skipping to di!erent subjects. (e) Pace yourself; work on assignments before
they are due to have enough time to do your best. (f ) Exercise regularly to reduce
stress.
196 D. Wachelka, R.C. Katz / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 191}198

The eighth and "nal session involved training in test taking (Sapp, 1993). Students
were encouraged to arrive early for the test so they had time to relax. They were told
to listen carefully to instructions given by the teacher, look over the entire test before
starting, develop a time plan for taking the test, read the instructions carefully, and jot
down memory aides they might need and also might forget. They were taught speci"c
strategies for taking multiple choice, true}false, and machine-graded tests. For
example, during a true}false test, they were encouraged to look for `quali"era words,
such as always, never, sometimes, and rarely. In doing so, they were told how these
words make the statement true or false. They were also told that `absolutea quali"ers
such as never or always usually indicate a false statement. Finally, they were given
guidelines for guessing, and speci"c instructions for taking open book, short answer,
and essay examinations.

3. Results

Mean pre- and post-test scores for each of the three dependent measures are shown
in Table 1. These data were analyzed separately using a 2×2 split-plot analysis of
variance. On all three measures, signi"cant group, trial, and interaction e!ects were
found. The e!ect of interest was the group by trial interaction. As predicted, the three
signi"cant interaction e!ects showed that the two groups did not di!er from each
other on any of the measures at pretest, but by posttest only the treatment group
improved. The F values (1, 25) for the interaction e!ect on the TAI, Survey of Study
Habits and Attitudes, and Coopersmith were 30.75, p(0.0001, 5.08, p"0.03, and
5.80, p"0.02, respectively.
Anecdotal evaluations of the treatment were obtained from students at the comple-
tion of the study during a debrie"ng interview. Their reports indicated that the
treatment was perceived as helpful in reducing apprehension about tests, building
con"dence about schoolwork, and improving attitudes toward school. Of the treat-
ment methods used, relaxation training was rated as being most helpful.

Table 1
Mean scores and standard deviations for the treated and control group across the three
dependent measures. The group by trial interaction for all three measures was signi"cant.
The p values are shown in the table

Measure Pretest Posttest F

Test Anxiety Inventory

Treated group 59.82 (11.25) 35.91 (8.76) p(0.0001
Control group 58.56 (9.87) 54.31 (10.81)
Survey of Study Habits and A ttitudes
Treated group 71.00 (25.37) 96.64 (29.54) p"0.03
Control group 72.00 (27.34) 76.25 (28.02)
Coopersmith Self-Esteem School Scale
Treated group 3.18 (1.6) 5.09 (1.38) p"0.02
Control group 3.20 (1.90) 3.67 (1.72)
 D. Wachelka, R.C. Katz / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 191}198 197

4. Discussion

Results showed that a cognitive-behavioral treatment using relaxation training,

guided imagery, self-instructional training, and study skills training reduced test
anxiety in high school and college students with LD. Improvements also occurred in
the students' academic self-esteem and self-rated e!ectiveness of their study skills.
Although similar results were reported by other investigators (Gonzales, 1995; Sapp,
1994; Wilson & Rotter, 1986), these are the "rst results of their kind from high school
and college students with LD. The results are encouraging because (a) signi"cant
improvements occurred on all three measures * i.e., they were not speci"c to reducing
anxiety alone; (b) improvements occurred following a relatively brief (8 week) treat-
ment; and (c) there was little attrition (15%) in the treated students despite the fact
that there were no penalties for dropping out, no special incentives for participating,
and the students were kept busy doing `homeworka.
This was an integrative treatment and thus we cannot say if any one method, or all
of them combined, was needed to produce the results we obtained. Likewise, the
underlying mechanisms responsible for the improvement remain to be identi"ed. One
possibility is that the treated students learned new skills that improved their study
habits, increased their con"dence, and helped them cope better with anxiety at an
emotional and cognitive level. It is also possible that the students responded to
experimental demand characteristics, expectations of improvement, or that they were
motivated to study harder which increased their con"dence that they could do well.
Of all the treatment methods used, the students rated relaxation training as being
most helpful to them. This anecdotal "nding suggests that improvements may have
occurred because students were coping di!erently in test situations by relaxing, or
because they felt less distressed and more empowered just knowing what to do,
whether or not they actually used the relaxation procedure. All of these are plausible
explanations. Which interpretation is most plausible is best determined by replication
studies with appropriate control groups that systematically dismantle the components
of the intervention in order to determine their e!ects.
It should also be noted we did not do a follow-up to assess the maintenance of
improvement, nor did we look at the e!ects of treatment on the students' actual test
performance or their grade point average. While these questions can be addressed in
future research, there is evidence (Gonzales, 1995; Algaze, 1995) that anxiety reduction
procedures similar to ours were e!ective in improving the grades of test anxious
college students. What we can conclude from this study is that only the students who
received treatment improved; they improved fairly quickly, and their improvement
was noted along several dimensions. While the results are preliminary, they o!er hope
to test anxious high school and college students with learning disabilities.

Acknowledgements

This study is based on Donald Wachelka's master's thesis that was done under the
direction of the second author. Appreciation is extended to committee members
198 D. Wachelka, R.C. Katz / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 191}198

Drs. Kenneth Beauchamp and Gary Howells, and to Ann Muskal, Melody Tennant,
and Lynn Swanson for the help they provided.

References

Algaze, B. (1995). Cognitive therapy, study counseling, and systematic desensitization in the treatment of
test anxiety. In C. Speilberger, & P. Vagg, Test anxiety: Theory assessment and treatment (pp. 133}152).
Washington DC: Taylor & Francis Publishers.
Ball, S. (1995). Anxiety and test performance. In C. Speilberger, & P. Vagg, Test anxiety: Theory assessment
and treatment (pp. 107}113). Washington DC: Taylor & Francis Publishers.
Brown, W., & Holtzman, W. (1967). SSHA manual: Survey of study habits and attitudes. New York: The
Psychological Corporation.
Bryan, J., Sonnefeld, L., & Grabowski, B. (1983). The relationship between fear of failure and learning
disabilities. Learning Disability Quarterly, 6, 217}222.
Coopersmith, S. (1981). Coopersmith Inventory. Palo Alto: Consulting Psychologists Press.
Davis, M., Eshelman, E., & McKay, M. (1996). The relaxation and stress reduction workbook. Oakland, CA:
New Harbinger Press.
Dusek, J. (1980). The development of test anxiety in children. In I. Sarason, Test anxiety: Theory, research,
and applications (pp. 87}110). Hillsdale, NJ: Lawrence Erlbaum Associates Publishers.
Ellis, B. (1985). Becoming a master student. Rapid City, SD: College Survival, Inc.
Goldfried, M., & Davison, G. (1994). Clinical behavior therapy. New York: Wiley-Interscience.
Gonzales, H. (1995). Systematic desensitization, study skills counseling, and anxiety coping training in the
treatment of test anxiety. In C. Speilberger, & P. Vagg, Test anxiety: Theory assessment and treatment
(pp. 107}116). Washington DC: Taylor & Francis Publishers.
Nicaise, M. (1995). Treating test anxiety. Teacher Education and Practice, 11, 65}81.
Rizzo, J., & Zabel, R. (1988). Educating children and adolescents with behavior disorders: An integrative
approach. Boston: Allyn & Bacon.
Sapp, M. (1993). Test anxiety: Applied research, assessment, and treatment interventions. Lanham, MD:
University Press of America.
Sapp, M. (1994). Cognitive-behavioral counseling: Applications for African}American middle school
students who are academically at-risk. Journal of Instructional Psychology, 21, 161}171.
Sapp, M., & Farrell, W. (1994). Cognitive-behavioral interventions: Applications for academically at-risk
and special education students. Preventing School Failure, 38, 16}24.
Scruggs, T., & Mastropieri, M. (1988). Are learning disabled students `test-wisea?: A review of recent
research. Learning Disabilities Focus, 3, 87}97.
Speilberger, C. (1980). Test anxiety inventory. Palo Alto, CA: Consulting Psychologist Press, Inc.
Strumph, J., & Fodor, I. (1993). The treatment of test anxiety in elementary school-age children: Review and
recommendations. Child & Family Behavior Therapy, 15, 19}42.
Surwit, R. (1983). Progressive muscle relaxation (audio cassette recording). North Carolina: Duke University.
Swanson, S., & Howell, C. (1996). Test anxiety in adolescents with learning disabilities and behavior
disorders. Exceptional Children, 62, 389}397.
Tobias, S. (1979). Anxiety research in educational psychology. Journal of Educational Psychology, 71,
573}582.
Vagg, P., & Speilberger, C. (1995). Treatment of test anxiety: Application of the transactional model. In
C. Speilberger, & P. Vagg, Test anxiety: Theory assessment and treatment (pp. 197}216). Washington DC:
Taylor & Francis Publishers.
Wessel, I., & Mersch, P. (1994). A cognitive-behavioural group treatment for test anxious adolescents.
Anxiety, Stress, and Coping, 7, 149}160.
Wilson, N., & Rotter, J. (1986). Anxiety management training and study skills counseling for students on
self-esteem and test anxiety and performance. The School Counselor, 4, 18}31.